[Link].

org/Langmuir Article

Self-Assembled Surfactant-Templated Synthesis of Porous Hollow

Silica Nanoparticles: Mechanism of Formation and Feasibility of
Post-Synthesis Nanoencapsulation
Vinicius Bueno and Subhasis Ghoshal*
Cite This: Langmuir 2020, 36, 14633−14643 Read Online

ACCESS Metrics & More Article Recommendations *

sı Supporting Information
Downloaded via AUCKLAND UNIV OF TECHNOLOGY on December 15, 2020 at 11:59:32 (UTC).
See [Link] for options on how to legitimately share published articles.

ABSTRACT: SiO2 is bioinert and highly functionalizable, thus

making it a very attractive material for nanotechnology applications
such as drug delivery and nanoencapsulation of pesticides. Herein,
we synthesized porous hollow SiO2 nanoparticles (PHSNs) by
using cetyltrimethylammonium bromide (CTAB) and Pluronic
P123 as the structure-directing agents. The porosity and
hollowness of the SiO2 structure allow for the protective and
high-density loading of molecules of interest inside the nanoshell.
We demonstrate here that loading can be achieved post-synthesis
through the pores of the PHSNs. The PHSNs are monodisperse
with a mean diameter of 258 nm and a specific surface area of 287
m2 g−1. The mechanism of formation of the PHSNs was
investigated using 1-D and 2-D solid-state nuclear magnetic resonance (SS-NMR) and Fourier-transform infrared spectroscopy
(FTIR). The data suggest that CTAB and Pluronic P123 interact, forming a hydrophobic spherical hollow cage that serves as a
template for the porous hollow structure. After synthesis, the surfactants were removed by calcination at 550 °C and the PHSNs
were added to an Fe3+ solution followed by addition of the reductant NaBH4 to the suspension, which led to the formation of Fe(0)
NPs both on the PHSNs and inside the hollow shell, as confirmed by transmission electron microscopy imaging. The imaging of the
formation of Fe(0) NPs inside the hollow shell provides direct evidence of transport of solute molecules across the shell and their
reactions within the PHSNs, making it a versatile nanocarrier and nanoreactor.

■ INTRODUCTION
The development of different methods to synthesize
potential, and thus are a promising encapsulation medium for
pesticides, nutrients, or other growth factors in agriculture.6
An alternative structure to MSNs that offers many of the
mesoporous SiO2 nanoparticles (MSNs) has generated a lot
advantages, but also a higher internal loading volume for active
of interest from the research community due to the promising
ingredient molecules, is the porous hollow SiO2 nanoparticle
application of MSNs as a nanocarrier system. SiO2 is an earth-
(PHSN). The higher internal volume can provide high local
abundant, bioinert compound and thus can be safely used for
concentrations of active ingredients and the potential for
many in vivo applications, particularly in medicine and
significantly higher loading per unit mass of silica compared to
agriculture.1 In medicine, for example, MSNs have been tested
MSNs. Furthermore, PHSNs can be used as a delivery agent
for targeted drug delivery,2 where drugs are loaded into the
that provides the slow release of active ingredient to the
SiO2 matrix and inoculated in HeLa cells, an immortal human
exterior over a more extended period. PHSNs are being tested
cell line. Regli et al.3 encapsulated ibuprofen and silicon
as drug delivery systems for cancer treatment for their slow
nanocrystals inside an MSN to create a platform for combined
release properties.7 The shell can also protect active
drug delivery and bioimaging. The high specific surface area of ingredients from the influence of certain environmental
MSNs (greater than 1000 m2 g−1) and the ability to tune pore conditions. Li et al.8 demonstrated that the PHSNs provided
sizes and to functionalize the surface with different UV shielding of the pesticide avermectin.
biomolecules enable the application of these nanoparticles
for targeted binding to specific cellular sites.4 In agriculture,
pesticides have been loaded in MSN nanocarriers and applied Received: August 23, 2020
to plants to assess whether the pesticides in nanoformulations Revised: November 4, 2020
either have higher uptake and longevity or offer greater Published: November 23, 2020
protection than pesticides applied with the traditional methods
without nanocarriers.5 Furthermore, inorganic compounds
such as silica are micronutrients for plants, have low toxicity

© 2020 American Chemical Society [Link]

14633 Langmuir 2020, 36, 14633−14643
Langmuir [Link]/Langmuir Article

A common approach for synthesis of PHSNs is to grow the of the transport of the target molecules into the hollow core of
silica matrix around hard-template materials, such as the PHSN. Slow release rates are often cited as evidence of
polystyrene beads9 or resorcinol-formaldehyde resin nano- loading inside nanocarriers, but release rates of compounds do
spheres,10 which are further removed by calcination or solvent not prove that the compounds were contained inside the
extraction to create a hollow-core nanoshell. This is a lengthy nanoparticle core because slow release can occur from the
procedure and can cause contamination in the PHSNs. The compounds adsorbed on the nanoparticle pore surfaces.
hard-template synthesis also requires the functionalization of We address this major knowledge gap in nanoencapsulation
the template to allow the deposition of the silica coating.11 by imaging internalized compounds post-synthesis in the
This requirement is eliminated with soft-template methods, PHSNs. Two different compounds (FeCl3·6H2O and NaBH4)
which involve the use of one or more surfactants associated were added sequentially to the PHSNs. The formation of
with a swelling agent, usually an oil phase. In the work of Fe(0) inside the PHSNs, as shown by TEM imaging, suggests
Wibowo et al.,12 Miglyol 812, a nontoxic oil used for that Fe3+ and BH4− migrated to the hollow core and reacted to
pharmaceutical formulations, was used along with a custom- form Fe(0) NPs. This post-synthesis encapsulation process
ized polypeptide to form a positively charged nanoemulsion, could suggest that not only transport of mass between the
which, in turn, attracted the silica precursor to the water−oil interior and exterior of the nanoshell does happen but also that
interface. reactions could take place within the PHSNs. Having
Herein, we report the synthesis protocol to produce PHSNs compounds internalized in a rigid porous structure such as a
using CTAB and Pluronic P123 only, without the use of a solid PHSN post-synthesis could ensure application of the nano-
or liquid (oil) phase as a template, thus reducing the excessive carriers with maximum loading and without the need for
use of organic solvents and the necessity to modify hard- subsequent steps (e.g., calcination or acid washing) that could
template structures. CTAB was selected because it is a well- otherwise damage the loaded compounds. Iron nanoparticles
known pore-directing agent for the synthesis of mesoporous encapsulated in silica have promising applications in power
silica nanoparticles since the introduction of MCM-41 in transformers, magnetic recording heads, microwave, and
2001.13 CTAB self-assembles, forming an ordered hexagonal magnetic sensing.28
mesophase micelle,14 which serves as a template for a It should be noted that encapsulating compounds during
symmetrical distribution (in terms of both size and structure) synthesis, as opposed to post-synthesis, may present some
of pores in the silica matrix after calcination. Pluronic P123 drawbacks: (1) if the shell is not porous, the encapsulated
with its spherical micellar structure comprised of hydrophobic material will not be released unless the nanoparticle is
core (polypropylene oxide) and hydrophilic corona (poly- disassembled, and (2) if the material has its pores filled by
ethylene oxide)15 served as the agent causing the formation of surfactants due to the synthesis, a subsequent step to open up
the hollow-core template for the PHSNs. these pores may be required, such as calcination and acid
The morphology and mechanism of formation of the extraction. Both calcination and acid extraction, however, may
PHSNs were studied using microscopy and spectroscopy risk destroying the encapsulated materials if they do not
techniques. Specifically, transmission electron microscopy possess heat and/or acid-resistant properties.
(TEM) was performed to assess the PHSN shape and size,
and nitrogen adsorption/desorption experiments were per-
formed to characterize surface area and pore size distribution.
■ EXPERIMENTAL SECTION
Materials. Tetraethyl orthosilicate (Si(OC2H5)4, TEOS; reagent
Solid-state nuclear magnetic resonance (NMR) and Fourier- grade, 98%), ammonium hydroxide solution (NH4OH, 28.0−30.0%
transform infrared spectroscopy (FTIR) analyses were NH3 basis), hexadecyltrimethylammonium bromide (C19H42BrN,
performed to identify bonds and the chemical environment CTAB), symmetric triblock copolymer Pluronic P123, and sodium
of different elements during the synthesis toward developing borohydride (NaBH4, 99.99%) were purchased from Sigma-Aldrich.
Deionized (DI) water ASTM type 1 was purchased from Thermo
an understanding of the mechanism of formation of the Fisher Scientific. Ethyl alcohol (anhydrous, 100%) was purchased
PHSNs. from Commercial Alcohols (Canada). Iron (III) chloride (FeCl3·
Although alternative micellar soft-templated synthesis has 6H2O, 99.9%) was purchased from Acros Organics.
been shown with a range of different surfactants,16−18 they Synthesis of the PHSNs. In a single-neck, 500 mL, round-bottom
produce PHSNs with very different characteristics in terms of flask, DI water, ethanol (33%, v/v), ammonium hydroxide solution
particle size, pore size distribution, and shell thickness. Herein, (6.7%, v/v), CTAB (3.6 mM), and Pluronic P123 (0.7 mM) were
we used CTAB and Pluronic P123 to produce PHSNs with added sequentially and stirred under 1000 rpm for 1 h until the
unique characteristics to be used in diverse downstream reagents were completely dissolved. Then, TEOS (45 mM) was
added dropwise at a rate of 0.75 mL/min. The solution acquired a
applications, such as the encapsulation of organic compounds
milky color after 2 min and remained under magnetic stirring (1000
or other nanoparticles. In addition, we characterized in detail rpm) for 5 h. Next, the solution was dried overnight under 80 °C. To
the mechanism of formation that dictates the structure of this remove the surfactants, the PHSN dried powder was calcined at 550
PHSN. Some other studies have used solely CTAB, solely °C for 5 h. Table S1 has the exact volumes and weight of each
Pluronic P123, or a combination of both to produce chemical used in a standard batch.
MSNs19−21 but not PHSNs. Furthermore, the previous studies Characterization of the PHSNs. The frequency distribution of
did not provide evidence for whether the particles had porous the hydrodynamic diameters of PHSNs was obtained from dynamic
shells. light scattering (DLS) using a Malvern Zetasizer Nano ZS (Malvern
Encapsulating functional biomolecules inside nanocarriers Instruments, U.K.). DLS analysis was conducted using PHSN
suspensions in DI water with a concentration of 100 ppm. The
for efficient transport and uptake at biological target locations morphology and thickness of the PHSN shell were determined by
is an emerging application of nanotechnology in medicine and transmission electron microscopy (TEM) using a Philips model
agriculture.22,23 However, although many studies24−27 report CM200 transmission electron microscope at an acceleration voltage of
encapsulation of drugs, pesticides, and biomolecules post- 200 kV. For TEM analysis, 10 μL of 100 ppm PHSN suspension was
synthesis, they do not provide direct evidence (e.g., imaging) placed on lacey carbon-coated grids and air-dried. The surface area

14634 [Link]
Langmuir 2020, 36, 14633−14643
Langmuir [Link]/Langmuir Article

Figure 1. The upper images represent (a) the particle size distribution of calcined PHSN obtained from DLS and (b) the particle size distribution
(N = 100) of calcined PHSN from TEM. The lower figures represent (c) the TEM image of calcined PHSNs, (d) the nitrogen sorption/desorption
isotherm type IV profile, and (d, inset) the pore size distribution of calcined PHSNs obtained from BET and BJH, respectively.

and pore size distributions were determined by nitrogen adsorption/ and NaBH4. All the steps were carried out inside an anaerobic
desorption experiments in a Quantachrome Autosorb-1 (Quantach- chamber containing an atmosphere of N2:H2 at a ratio of 99:1 or in
rome GmbH & Co., Netherlands) using the Brunauer−Emmett− sealed vials assembled inside the chamber. First, FeCl3·6H2O (600
Teller (BET) and Barrett−Joyner−Halenda (BJH) approaches, mg) was solubilized in a mixture of DI water (21 mL) and ethanol (9
respectively. mL) previously purged with nitrogen for 40 min. Second, 400 mg of
Understanding the Synthesis Mechanism of Formation PHSNs was suspended in the FeCl3 solution in a sealed vial and
with SS-NMR and FTIR. Solid-state NMR experiments were sonicated in a water bath at 37 kHz for 60 min (Fisherbrand 112xx
performed on a Varian VNMRS with a wide-bore 9.4 T magnet Series, Thermo Fisher Scientific, operated at 100% power). The
and Varian T3 double-resonance 4 mm probe. A set of one- suspension was then centrifuged at 8500 rpm for 10 min and the
dimensional (1-D) and two-dimensional (2-D) NMR was performed, supernatant was discarded. Next, the pellet was washed with degassed
including 1-D single-pulse magic angle spinning (MAS) 1H, 1-D DI water to remove the non-internalized ferrous ions. A second
cross-polarization (CP) MAS 13C (3000 number of scans), 1-D CP- solution was prepared by dissolving NaBH4 (200 mg) in degassed DI
MAS 29Si (512 number of scans), 2-D 1H−1H exchange MAS, 2-D water (6 mL). Then, 2 mL of the NaBH4 solution was added
1
H−13C heteronuclear correlation (HETCOR), and 2-D 1H−29Si dropwise over the FeCl3@PHSN pellet. The pellet was allowed to dry
HETCOR. The resonance frequencies for 1H, 13C, and 29Si were inside the chamber.
399.8, 100.5, and 79.4 MHz, respectively. The spin rates were 10 kHz. The PHSNs with the FeNPs were imaged and analyzed with TEM
The recycle delays were 4 s, except to compute the relaxation times T1 and energy-dispersive X-ray (EDX) spectrum using a Philips model
and T1ρ when it was 120 s. Total times for the 2-D experiments were CM200 transmission electron microscope at an acceleration voltage of
24, 18, and 1.5 h for 1H−13C HETCOR, 1H−29Si HETCOR, and 200 kV by placing the dried powder directly on a lacey carbon-coated
TEM grid.

■
1
H−1H exchange, respectively. Acquisition times were 15 ms for 13C,
20 ms for 29Si, 200 ms for 1H, and 2 s for 1H−1H exchange. CP
contact times were 1500 μs for 13C and 1000 μs for 29Si HETCOR. RESULTS AND DISCUSSION
The mixing times for 1H−1H exchange were 20 and 100 ms. The PHSN Characterization. The Z-average hydrodynamic
spectra were obtained for calcined PHSN and precalcined PHSN. diameter measured by DLS was 258 nm (Figure 1a). The size
FTIR spectra for pure CTAB, Pluronic P123, and PHSN pre- and distribution showed a monodisperse population of particles
post-calcination were obtained with a Spectrum II (Perkin Elmer) with a polydispersity index (PdI) of 0.187. Similar particle size
spectrometer with a single bounce diamond crystal. Spectra were distribution was obtained from TEM (Figure 1b) with an
recorded in the range from 4000 to 400 cm−1 at a resolution of 1
average diameter of 253 nm (N = 100). The TEM image
cm−1. Approximately 70 mg of each sample was deposited on the
diamond crystal for analysis. (Figure 1c) showed synthesized PHSNs as spherical shells,
Assessment of Encapsulation in PHSN Post-Synthesis and with most of the particles having a shell thickness ranging from
Calcination. To obtain direct visual proof if the void in the PHSNs 22 to 38 nm and the core diameter ranging from 130 to 151
can contain mass that is supplied externally post-synthesis, we created nm. The size of the pores is not clear from the TEM images;
iron nanoparticles in the presence of PHSNs by adding FeCl3·6H2O therefore, nitrogen adsorption−desorption experiments were
14635 [Link]
Langmuir 2020, 36, 14633−14643
Langmuir [Link]/Langmuir Article

Figure 2. The upper images are the SS-NMR spectra of (a) as-synthesized PHSNs and (b) calcined PHSNs. The lower images are schematic
representations of (c) Q3 and (d) Q4 Si.

performed to identify and quantify the porosity in the SiO2 Interactions and Bonds between Pluronic P123,
shell. The surface area obtained from the BET method was 287 CTAB, and SiO2 with the SS-NMR Spectra. 1-D CP-MAS
m2 g−1 and, as shown in Figure 1d, the nitrogen sorption/ 29
Si (Figure 2a), 1-D CP-MAS 1H (Figure 3a), 1-D CP-MAS
13
desorption profile exhibited characteristic type IV isotherm, C (Figure 3b), 2-D 1H−29Si HETCOR (Figure 3c), 2-D
which is expected in the presence of cylindrical micropores.29 1
H−13C HETCOR (Figure 3d), and 2-D 1H−1H exchange
The pore size distribution obtained from the BJH approach MAS (Figure 4) SS-NMR spectra were obtained for
confirmed the presence of micropores, ranging between 1.5 precalcined PHSN (pcal-PHSN), and 1-D CP-MAS 29Si
and 2 nm (Figure 1d, inset). The information about the (Figure 2b) was obtained for calcined PHSN (PHSN). Note
available volume inside the shell and the pore size distribution that both Q3 Si (SiOH) and Q4 Si (SiOSi)
helps us understand what kind of compounds could be were present in the pcal-PHSN (Figure 2a) and only Q4 Si in
internalized and confined as cargo within the PHSNs. the PHSN (Figure 2b). Q4 Si represents Si atoms bound to O
Interestingly, in this study, the size of the nanoparticles did atoms, which in their turn bind to other Si atoms (Figure 2d).
not change by varying the concentration of surfactants. In Q3 Si, however, only three oxygens are bound to other Si
However, the structure itself changed significantly when the atoms, whereas the fourth oxygen is associated with H (Figure
ratio between CTAB and Pluronic P123 varied as seen in the 2c). Analysis of both the SS-NMR and FTIR spectra
TEM images of Figure S1. While the optimal ratio of (discussed in the following section) suggests that Q3 Si and
CTAB:Pluronic P123 to synthesize porous hollow silica Q4 Si were present in the conjugate between a Pluronic-CTA+
nanoparticles was found to be 0.35:1 wt %, in cases with complex and the SiO2 precursor in the precalcined system. The
excess CTAB, the nanoparticle no longer featured a shell but presence of Q3 and Q4 Si in pcal-PHSN and the absence of Q3
rather formed a solid-porous silica nanoparticle, similar to an Si in PHSN suggest a shift in the chemical environment around
MSN (Figure S1a). When, however, Pluronic P123 was in the Si atom as a result of calcination.
excess, the nanoparticle featured a larger hollow core and a In the 1-D CP-MAS 1H spectrum of pcal-PHSN (Figure 3a),
much thinner SiO2 shell (Figure S1b). It is interesting to note we can identify the peaks for the protons in Pluronic, CTAB,
that in some studies using different surfactants, particle size and Q3 Si. For Pluronic P123, the methyl (CH3) group in the
varied with different surfactant concentrations but with no PPO block is assigned at 1.27 ppm, the methylene (CH2)
effect on the structure itself. For instance, in the work of group in the PPO block is assigned at 3.82 ppm, the methine
Nakashima et al.30 using polyacrylic acid, they obtained group bound to oxygen in the PPO block is assigned at 3.51
particles between 10 and 20 nm. In the work of Ikari et al.31 ppm, and both methylene groups bound to oxygen in the PEO
using hexadecyltrimethylammonium chloride (CTAC) and block are assigned at 3.68 ppm. These observations are
Pluronic F127, they obtained porous particles from 50 to 150 consistent with the work of Bae and Han.33 The peak at 1.27
nm. Last, in the work of Lv et al.32 using CTAC and ppm and the shoulder at 1 ppm were attributed to an overlap
triethanolamine (TEA), they obtained porous particles of signals from the methylene groups of CTAB and the methyl
between 20 and 110 nm. (CH3) group in the PPO block.34 The methyl groups of the
14636 [Link]
Langmuir 2020, 36, 14633−14643
Langmuir [Link]/Langmuir Article

Figure 3. The upper image illustrates Pluronic P123 and CTAB structures with protons labeled from a to h. The lower images represent the pcal-
PHSN SS-NMR spectra of (a) 1-D CP-MAS 1H, (b) 1-D CP-MAS 13C, (c) 2-D 1H−29Si HETCOR, and (d) 2-D 1H−13C HETCOR.

Figure 4. PHSN SS-NMR spectra of 2-D 1H−1H exchange MAS with mixing times of (a) 20 ms and (b) 100 ms.

hydrophilic head of CTAB are assigned in the same region and that the hydrophilic head of CTAB aligns with the less
could explain the shoulder at 1.46 ppm. Once again, there may hydrophobic moieties (PEO chains) and away from the
be an overlap between the resonance for −CH2−N(CH3)3 hydrophobic core of Pluronic (the PPO chain). The negatively
protons of the hydrophilic head of CTAB and the methylene charged hydrolyzed TEOS (SiO2 precursor) is attracted by the
and methine groups in Pluronic. This could potentially suggest positively charged Pluronic-CTAB structure, as hypothesized
14637 [Link]
Langmuir 2020, 36, 14633−14643
Langmuir [Link]/Langmuir Article

above, provided by the CTA+ ions, which are located in the

conjugate−water interface, whereas the PPO hydrophobic core
is confined in the core. The broad peak at 5.37 ppm is assigned
to the protons in Q3 Si, which is confirmed by the HETCOR
1
H−29Si spectra (Figure 3c). There, the resonance is observed
exactly in the Q3 Si peak and the peak at 5.37 ppm. In the CP-
MAS 13C (Figure 3b) spectrum, the peaks can be assigned for
the carbons in CTAB and Pluronic molecules with the support
of the HETCOR 1H−13C spectra (Figure 3d). The peak at 30
ppm, which had the highest density of protons, is assigned to
the carbon in the methyl groups for both Pluronic and CTAB.
The peak at 53.6 ppm is assigned to the methine group in the
PPO block of Pluronic and the peak at 70.6 ppm is right in
between the methylene groups in PPO and PEO blocks.
The 2-D exchange 1H−1H correlation spectra indicate how
neighboring protons interact and their chemical environment
at mixing times of 20 ms (Figure 4a) and 100 ms (Figure 4b).
As the experiment proceeds from 20 to 100 ms, we can observe
an interaction among the protons in the Q3 Si and protons
from the hydrocarbons, suggesting that they are sufficiently
close. The relaxation times (T1 and T1ρ), as shown in Table S2,
represent one common T1 (values ranging from 0.442 and
0.571 s with a standard deviation of 9%) and multiple T1ρ
(values ranging from 0.004 to 0.044 s with a standard deviation
of 55%), which suggests that the average distance values
between the functional groups are between 5 and 50 nm.35
The spread of the proton resonance in the HETCOR 1H−29Si
spectra (Figure 4a,b) toward the methylene groups and the
presence of Q3 Si in the PHSN could possibly suggest that the
Si−O groups are sufficiently close to the methylene groups
only present in the PEO chains of Pluronic.
Overall, the SS-NMR spectra observations show the
following: (i) a chemical shift around the Si atom (possibly
caused by the interaction between the negatively charged SiO2
colloids and the positively charged CTA+ cations from the
Pluronic-CTA+ complex) and (ii) an overlap between the
methylene groups of CTAB and methyl groups in the PPO
block of Pluronic P123 (possibly indicating that the CTAB
hydrophilic head and PEO chains of Pluronic in the
surfactant−water interface, whereas the PPO hydrophobic
chains are confined in the core).
These observations were only possible because SS-NMR
techniques track Si, H, and C atoms individually, while less
specific techniques, such as FTIR, is best used to define the Figure 5. FTIR spectra of the (a) 400−200 cm−1 region, (b) 1300−
presence or absence of functional groups. That is, although 1800 cm−1 region, and (c) 2750−3050 cm−1 region.
FTIR, a less specific technique, can be used complementarily
to support NMR observations, FTIR alone could not precisely (Figure 5a−c), all of which are present in the pcal-PHSN
explain what is happening with functional groups individually. spectrum and absent in the PHSN spectrum. These bands are
Martins et al.36 showed how combining FTIR and NMR assigned as below.
techniques permitted a comprehensive investigation of the The band at 955 cm−1 present in the pcal-PHSN possibly
chemical environment around silicate domains. Below, we indicates the presence of Si−OH bond (Figure 5a). The
describe the FTIR spectra and indicate how they are consistent absence of the same band in the PHSN supports the
with the observations from the SS-NMR. observations in the SS-NMR and possibly suggests that Q3
Using FTIR to Complement the Information Obtained Si (Figure 2c) content is transformed into Q4 Si (Figure 2d),
with SS-NMR. FTIR spectra were obtained for pure CTAB, eliminating the band assigned to the Si−OH bond. This could
pure Pluronic P123, precalcined PHSN (pcal-PHSN), and suggest a shift in the chemical environment around the Si atom
calcined PHSN (PHSN), as shown in Figure 5. Clear after the calcination and, as discussed in the 1-D CP-MAS 29Si
differences between the pcal-PHSN and PHSN spectra could spectra, this shift could be attributed to the interaction
be observed (Figure 5a−c), which could be related to the fact between the SiO2 precursor and the CTA+ ions of the
that the precalcined particles still have surfactants and residual Pluronic-CTA+ complex.
ammonia (from NH4OH) associated with it from the The 3000−2800 cm−1 region shows characteristic bands for
synthesis. More specifically, these differences are noted in CTAB (C−H stretch of sp3 hybridized carbon −CH3 at 2855
the 1000−900, 1500−1300, and 3000−2800 cm−1 regions cm−1 and C−H stretch of sp3 hybridized carbon −CH2 at 2925
14638 [Link]
Langmuir 2020, 36, 14633−14643
Langmuir [Link]/Langmuir Article

Figure 6. Schematic representation of the formation of CTAB-Pluronic emulsion, followed by its silanization with TEOS and calcination to make
the PHSNs ready for downstream applications.

cm−1) and for Pluronic P123 (symmetric C−H stretch of Differences between pure materials and pcal-PHSN spectra
−CH2 broad peak from 2893 to 2865 cm−1) (Figure 5b). Even can also be observed in the 1500−1300 cm−1 region (Figure
though these characteristic bands are present in the pcal-PHSN 5b). The Pluronic peak at 1344 cm−1 (CH2 wag) slightly
and absent in PHSN, some of them are shifted toward high shifted to higher energy levels. The region between 1600 and
energy levels and/or are broader than the ones observed in the 1650 cm−1 is only present in pcal-PHSN due to the N−H
pure surfactants. These changes could possibly suggest that the bending39 in residual ammonia in the system from the
chemical environment of these bonds is different in the pure synthesis process (Figure 5b).
molecules from when they are associated with the silica Mechanisms of Formation and Structure of PHSN.
nanoparticles. One change is the shift of the CTAB The TEM image, pore size distribution, SS-NMR spectra, and
characteristic bands from 2850 cm−1 (C−H stretch of FTIR spectra provide meaningful information about the shape
−CH3) and 2915 cm−1 (C−H stretch of −CH2) toward and morphology of the PHSN as well as the chemical
2855 and 2925 cm−1 (Figure 5c). This shift in the energy level environment and bonds involved in the structure. Combining
is consistent with the work of Poyraz et al.37 and with the all this information, we suggest the possible mechanism of
observations in the 1-D CP-MAS 29Si (Figure 2a,b). In fact, formation of the PHSN (Figure 6). From the FTIR spectra,
there are indications that Pluronic P123 formed a positively
they concluded that CTA+ ions from CTAB are crucial for the
charged complex with CTA+ ions, possibly attracting/
surfactant-SiO2 assembly because the CTA+ ions provide
anchoring SiO2 from the hydrolyzed TEOS to create the
positive charge to the Pluronic-CTA+ complex, thus attracting
SiO2 shell. Q4 Si (Figure 2d) is more hydrophobic and tends to
the silica precursor with a large negative zeta potential at be formed toward the core, while the more hydrophilic Q3 Si
neutral and basic pH.37,38 The Pluronic broad band from 2865 (Figure 2c) is located in the interface between the surfactant
to 2893 cm−1 (symmetric C−H stretch of −CH2) almost complex and aqueous phase, allowing condensation between
disappeared (Figure 5c). The Pluronic peak at 2933 cm−1 hydrolyzed TEOS to happen and thus growing the silica shell.
(asymmetric C−H stretch of −CH2) also disappeared or was The observation of the proximity among the PEO chains,
overlapped (Figure 5c). As intensity provides information CTAB, and the Q3 Si is also consistent with the FTIR and SS-
about concentration and the amount of CH2 groups in 850 mg NMR spectra. The PEO extremities of Pluronic P123 aligned
of Pluronic P123 is greater than the amount of CH2 groups in with each other, forming a hydrophobic core structure of PPO
300 mg of CTAB, the suppression or overlapping of bands in chains with outward projections comprised of PEO chains and
those regions of the pcal-PHSN spectra could indicate CTAB (Figure 6). CTAB deposited in a parallel manner to and
structural changes related to the CH2 groups in Pluronic in between the PEO chains of Pluronic. The hydrophobic tail
P123. As explained above for the CTA+ ions, these shifts in of CTAB was in proximity to the PPO hydrophobic region,
energy levels may be attributed to the formation of the whereas the hydrophilic head was pointed toward the PEO
Pluronic-CTA+ complex. hydrophilic terminus. This structure is consistent with the
14639 [Link]
Langmuir 2020, 36, 14633−14643
Langmuir [Link]/Langmuir Article

Figure 7. Fe(0)NP@PHSN (a−f) TEM images and (g) EDX spectrum of the marked site (*) in (a).

Pluronic-CTA+ complex model proposed by Poyraz et al.37 existence of Fe was confirmed with EDX (Figure 7g), which
Although Poyraz et al. demonstrated the formation of a showed peaks for Fe, indicating the presence of iron species,
Pluronic-CTA+ complex, this study shows for the first time that and those for Si and O, indicating the presence of SiO2; the Cu
a Pluronic-CTA+ complex can self-assemble in a configuration and C peaks are from the lacey TEM grids, which have a base
that allows the SiO2 precursor to form a spherical template for of copper and are coated with carbon; and last, the Na peak is
PHSN synthesis. The complexes formed by Pluronic P123 and presumably derived from unwashed sodium from NaBH4 that
CTAB in the study by Poyraz et al. led to the formation of remained in the system.
templates for silica nanorods. The hydrophobic core of the Previous studies24−27 have suggested the encapsulation of
nanoemulsion in this study corresponds to the template onto organic and inorganic compounds within the PHSNs post-
which the SiO2 precursor electrostatically attaches to form the synthesis, but whether the location and distribution of the
PHSN shell and the CTAB long hydrocarbon chains were encapsulated molecules are on or inside the hollow nanocages
responsible for the mesoscaled pores in the shell. After were not reported. In this study, we demonstrate through
calcination, the Pluronic P123 core and CTAB were removed direct observation that molecules could be internalized in the
from the system, making the SiO2 nanoshell hollow and PHSN core. Because a metal such as Fe has higher contrast in
porous. TEM compared to SiO2, it was a promising candidate for
Internalization of FeCl3/NaBH4 and Formation of Iron obtaining an image of an internalized mass inside the particle.
Nanoparticles inside the PHSNs. The PHSNs in the TEM In several PHSN, Fe(0) NPs were observed inside the silica
images (Figure 7a−f) were filled with darker smaller circles, shell, suggesting that the particles were formed inside the
which indicated the presence of Fe(0) nanoparticles. The PHSNs. This would not be possible if NaBH4 and FeCl3 did
14640 [Link]
Langmuir 2020, 36, 14633−14643
Langmuir

■
[Link]/Langmuir Article

not diffuse into the PHSNs. Some Fe(0) NPs were also CONCLUSIONS
observed to be attached on the PHSNs, which suggests that The range of applications of porous nanoparticles has been
some of Fe3+ did not stay internalized in the core. The TEM increasing exponentially in different areas such as biomedicine
images (Figure 7a−f) also show that only one particle is and agriculture. This study demonstrates a new, more atom-
present within each shell and that it can be explained by the efficient synthesis of PHSNs, which does not require a solid or
fact that once FeCl3, NaBH4, and H2O react, Fe(0) is formed, nonaqueous phase as a template for synthesis. NMR and FTIR
which precipitates. The precipitated Fe(0) colloids create a analyses suggest that Pluronic P123 and CTA+ ions assembled,
nucleation seed for the growth of Fe(0) particles; thus, the size forming a Pluronic-CTA+ complex. Thereafter, the positive
of the iron oxide particle inside the shell is determined by the charge provided by the CTAB hydrophilic head attracts the
amount of reagents reacting within the shell. Figure 7b,d silica precursor and facilitates the growth of the silica shell.
clearly shows differences in Fe(0) particle size inside the SiO2 Furthermore, although several studies claim to have
shell. The transport and loading of molecules would be internalized molecules within the PHSNs, there was no
controlled by the size of the PHSN, the porosity of the shell, evidence if the molecules were either indeed internalized or
and how the molecules interact with SiO2. adsorbed in the pore surfaces or whether they were adsorbed
Although TEM images are two-dimensional, we can observe on the external surface of the nanoshell. The internalization of
consistent features in the several images, which suggests that FeCl3 followed by the subsequent internalization of NaBH4
Fe(0) was indeed in the hollow core and not only on the resulted in the reduction of Fe3+ to Fe(0) that precipitated and
exterior surface. First, the Fe(0) particles were always started the nucleation/growth of Fe(0) NPs inside and on the
associated with the PHSNs and not once found unassociated shell. This experiment was a proof of concept that the shell was
with the PHSNs, suggesting that most of Fe3+ was loaded on or indeed porous hollow and allows transport of mass through it.
in the particles. Second, we can observe that when the Fe(0) Understanding how molecules are internalized within
particle was close to the walls of the PHSNs (Figure 7c,e), the nanocages can enhance nanotechnology applications in drug
circular shape was flattened out in the region in contact with delivery and agriculture.

■
the shell. This was more noticeable in Figure 7e, where the
four external Fe(0) particles are flattened in the region of ASSOCIATED CONTENT
contact with the particle, whereas the single Fe(0) particle
inside the shell kept its circular shape. This was also somewhat
* Supporting Information
sı
The Supporting Information is available free of charge at
noticeable in Figure 7c with the iron particles presumably
[Link]
inside the shell having a region flattened where it contacted the
silica shell. Shells can be seen around the Fe(0) nanoparticles, Comprehensive list of chemicals and the concentrations
particularly in Figure 7b−d, and these are typical iron oxide used for the synthesis of PHSNs, TEM images of the
layers resulting from anaerobic corrosion reactions with excess CTAB/Pluronic P123 effects on particle
water.40 Based on the TEM images alone, we have not morphology, relaxation times T1 and T1ρ from 1-D
found evidence that Fe(0) nanoparticles were formed inside CP-MAS 1H, and schematic representation of the
the pores of the SiO2 shell. It is possible that because the pores internalization of FeCl3 and NaBH4 within the PHSNs
are in the microporous range, the available space restricted the (PDF)

■
formation of a seed and the subsequent nanoparticle. Some
studies have shown the formation of gold nanoparticles inside AUTHOR INFORMATION
the pores of MSN, such as MCM-41.41 Corresponding Author
The internalization of FeCl3 and NaBH4 and the formation
Subhasis Ghoshal − Department of Civil Engineering, McGill
of Fe(0) both inside and on the particle could be explained as
University, Montreal, Quebec H3A 0C3, Canada;
follows. (i) When the calcined PHSN was immersed in the
[Link]/0000-0001-9968-6150; Phone: (1)-514-398-
FeCl3 solution, some of the Fe and Cl ions infiltrated the pores 6867; Email: [Link]@[Link]; Fax: (1)-514-
and filled the hollow core. (ii) After centrifugation, decanting 398-7361
of the supernatant, and rinsing of the PHSNs with DI water,
Fe3+ could be found primarily inside the PHSNs as the non- Author
internalized ions were washed away. However, after rinsing Vinicius Bueno − Department of Civil Engineering, McGill
with DI water, some of the ions inside the shell could possibly University, Montreal, Quebec H3A 0C3, Canada;
diffuse out of the PHSN or remain in the pores of PHSNs. It is [Link]/0000-0002-7136-6400
also possible that some Fe3+ was bound to silica by electrostatic
Complete contact information is available at:
interactions or complexation.42−44 (iii) Later, when the sodium
[Link]
borohydride solution was added to the FeCl3@PHSN pellet,
Na+ and BH4− ions also infiltrated the pores, diffused into the
Author Contributions
Fe3+-filled core, and reacted to form Fe(0) inside of the PHSN,
V.B. and S.G. designed the study and wrote the manuscript.
or in some cases on its outer surface. This imaging is a proof of
V.B. performed the experiments.
concept that molecules can be loaded either within the
nanoshell or attached onto its walls. It should be noted that Notes
The authors declare no competing financial interest.

■
encapsulation of iron oxide nanoparticles within nonporous
silica nanoshells has been reported by Su et al.45 However, in
that study, the iron oxide particles in the core were formed ACKNOWLEDGMENTS
through thermal treatment from iron salts that were The research was funded by the Natural Sciences and
encapsulated in the silica shell during its synthesis, rather Engineering Research Council of Canada (grant nos.
than the transport of dissolved iron through the silica shell. RGPIN-2016-05022 and STPGP 506450-17) and by the
14641 [Link]
Langmuir 2020, 36, 14633−14643
Langmuir [Link]/Langmuir Article

Fonds de Recherche du Québec−Nature et Technologies (18) Zhou, X.; Cheng, X.; Feng, W.; Qiu, K.; Chen, L.; Nie, W.; Yin,
(grant no. 286120). V.B. was supported in part by the McGill Z.; Mo, X.; Wang, H.; He, C. Synthesis of hollow mesoporous silica
Engineering Doctoral Award. We thank Julio Terra from nanoparticles with tunable shell thickness and pore size using
McGill University for his insightful comments and Robin Stein amphiphilic block copolymers as core templates. Dalton Trans.
2014, 43, 11834−11842.
from the McGill Chemistry Characterization facility for the
(19) Poyraz, A. S.; Dag, Ö . Role of organic and inorganic additives
assistance with SS-NMR analysis. on the assembly of CTAB-P123 and the morphology of mesoporous

■ REFERENCES
(1) Malone-Povolny, M. J.; Schoenfisch, M. H. Extended nitric
silica particles. J. Phys. Chem. C 2009, 113, 18596−18607.
(20) Wang, Y.; Song, H.; Zhang, H.; Liao, L.; Liu, N.; Chen, X.
Direct synthesis of flat cake-type ordered mesoporous carbon in a
double surfactant system of P123/CTAB. J. Mater. Chem. 2011, 21,
oxide-releasing polyurethanes via S-nitrosothiol-modified mesoporous
silica nanoparticles. ACS Appl. Mater. Interfaces 2019, 11, 12216− 5576−5579.
(21) Yang, J.; Chen, W.; Ran, X.; Wang, W.; Fan, J.; Zhang, W.-x.
12223.
(2) Noureddine, A.; Hjelvik, E. A.; Croissant, J. G.; Durfee, P. N.; Boric acid assisted formation of mesostructured silica: from hollow
Agola, J. O.; Brinker, C. J. Engineering of large-pore lipid-coated spheres to hierarchical assembly. RSC Adv. 2014, 4, 20069−20076.
mesoporous silica nanoparticles for dual cargo delivery to cancer cells. (22) Patra, J. K.; Das, G.; Fraceto, L. F.; Campos, E. V. R.;
J. Sol-Gel Sci. Technol. 2019, 89, 78−90. Rodriguez-Torres, M. d. P.; Acosta-Torres, L. S.; Diaz-Torres, L. A.;
(3) Regli, S.; Kelly, J. A.; Barnes, M. A.; Andrei, C. M.; Veinot, J. G. Grillo, R.; Swamy, M. K.; Sharma, S.; Habtemariam, S.; Shin, H.-S.
C. Mesoporous silica encapsulation of silicon nanocrystals: synthesis, Nano based drug delivery systems: recent developments and future
aqueous dispersibility and drug release. Mater. Lett. 2014, 115, 21−24. prospects. J. Nanobiotechnol. 2018, 16, 71.
(4) Manzano, M.; Vallet-Regí, M. Mesoporous silica nanoparticles (23) de Oliveira, J. L.; Campos, E. V. R.; Bakshi, M.; Abhilash, P. C.;
for drug delivery. Adv. Funct. Mater. 2020, 30, 1902634. Fraceto, L. F. Application of nanotechnology for the encapsulation of
(5) Djiwanti, S. R.; Kaushik, S. Nanopesticide: Future Application of botanical insecticides for sustainable agriculture: prospects and
Nanomaterials in Plant Protection. In Plant Nanobionics; Springer: promises. Biotechnol. Adv. 2014, 32, 1550−1561.
2019; pp. 255−298. (24) Fu, Z.; Li, L.; Wang, Y.; Chen, Q.; Zhao, F.; Dai, L.; Chen, Z.;
(6) Popat, A.; Liu, J.; Hu, Q.; Kennedy, M.; Peters, B.; Lu, G. Q. M.; Liu, D.; Guo, X. Direct preparation of drug-loaded mesoporous silica
Qiao, S. Z. Adsorption and release of biocides with mesoporous silica nanoparticles by sequential flash nanoprecipitation. Chem. Eng. J.
nanoparticles. Nanoscale 2012, 4, 970−975. 2020, 382, 122905.
(7) Cheng, K.; Zhang, Y.; Li, Y.; Gao, Z.; Chen, F.; Sun, K.; An, P.; (25) Jiang, X.; Ward, T. L.; Cheng, Y.-S.; Liu, J.; Brinker, C. J.
Sun, C.; Jiang, Y.; Sun, B. A novel pH-responsive hollow mesoporous Aerosol fabrication of hollow mesoporous silica nanoparticles and
silica nanoparticle (HMSN) system encapsulating doxorubicin encapsulation of L-methionine as a candidate drug cargo. Chem.
(DOX) and glucose oxidase (GOX) for potential cancer treatment. Commun. 2010, 46, 3019.
J. Mater. Chem. B 2019, 7, 3291−3302. (26) Liu, F.; Wen, L. X.; Li, Z.-Z.; Yu, W.; Sun, H.-Y.; Chen, J.-F.
(8) Li, Z.-Z.; Chen, J.-F.; Liu, F.; Liu, A.-Q.; Wang, Q.; Sun, H.-Y.; Porous hollow silica nanoparticles as controlled delivery system for
Wen, L.-X. Study of UV-shielding properties of novel porous hollow water-soluble pesticide. Mater. Res. Bull. 2006, 41, 2268.
silica nanoparticle carriers for avermectin. Pest Manage. Sci. 2007, 63, (27) Tan, S.; Wu, Q.; Wang, J.; Wang, Y.; Liu, X.; Sui, K.; Deng, X.;
241−246. Wang, H.; Wu, M. Dynamic self-assembly synthesis and controlled
(9) Qi, G.; Wang, Y.; Estevez, L.; Switzer, A. K.; Duan, X.; Yang, X.; release as drug vehicles of porous hollow silica nanoparticles.
Giannelis, E. P. Facile and scalable synthesis of monodispersed Microporous Mesoporous Mater. 2011, 142, 601.
spherical capsules with a mesoporous shell. Chem. Mater. 2010, 22, (28) Santra, S.; Tapec, R.; Theodoropoulou, N.; Dobson, J.; Hebard,
2693−2695. A.; Tan, W. Synthesis and characterization of silica-coated iron oxide
(10) Lu, T.; Zhu, Y.; Xu, J.; Ke, M.; Zhang, M.; Tan, C.; Fu, Z.; nanoparticles in microemulsion: the effect of nonionic surfactants.
Qian, H. Evaluation of the toxic response induced by azoxystrobin in Langmuir 2001, 17, 2900−2906.
the non-target green alga Chlorella pyrenoidosa. Environ. Pollut. 2018, (29) Huh, S.; Wiench, J. W.; Yoo, J.-C.; Pruski, M.; Lin, V. S.-Y.
234, 379−388. Organic functionalization and morphology control of mesoporous
(11) Nakashima, Y.; Takai, C.; Razavi-Khosroshahi, H.; Fuji, M. silicas via a co-condensation synthesis method. Chem. Mater. 2003,
Effects of cations on the size and silica shell microstructure of hollow 15, 4247−4256.
silica nanoparticles prepared using PAA/cation/NH4OH template. (30) Nakashima, Y.; Takai, C.; Razavi-Khosroshahi, H.; Fuji, M.
Colloids Surf., A 2020, 593, 124582. Influence of the PAA concentration on PAA/NH3 emulsion template
(12) Wibowo, D.; Zhao, C.-X.; Peters, B. C.; Middelberg, A. P. J. method for synthesizing hollow silica nanoparticles. Colloids Surf., A
Sustained release of fipronil insecticide in vitro and in vivo from 2018, 546, 301−306.
biocompatible silica nanocapsules. J. Agric. Food Chem. 2014, 62, (31) Ikari, K.; Suzuki, K.; Imai, H. Structural control of mesoporous
12504−12511. silica nanoparticles in a binary surfactant system. Langmuir 2006, 22,
(13) Vallet-Regi, M.; Rámila, A.; del Real, R. P.; Pérez-Pariente, J. A 802.
new property of MCM-41: drug delivery system. Chem. Mater. 2001, (32) Lv, X.; Zhang, L.; Xing, F.; Lin, H. Controlled synthesis of
13, 308−311. monodispersed mesoporous silica nanoparticles: Particle size tuning
(14) Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C.; and formation mechanism investigation. Microporous Mesoporous
Beck, J. S. Ordered mesoporous molecular sieves synthesized by a Mater. 2016, 225, 238.
liquid-crystal template mechanism. Nature 1992, 359, 710−712. (33) Bae, Y. K.; Han, O. H. Removal of copolymer template from
(15) Luo, H.; Jiang, K.; Liang, X.; Hua, C.; Li, Y.; Liu, H. Insights SBA-15 studied by 1H MAS NMR. Microporous Mesoporous Mater.
into Morphological Transition of Pluronic P123 Micelles as a 2007, 106, 304−307.
Function of Gallate. Colloids Surf., A 2019, 572, 221−229. (34) Trébosc, J.; Wiench, J. W.; Huh, S.; Lin, V. S.-Y.; Pruski, M.
(16) Khanal, A.; Inoue, Y.; Yada, M.; Nakashima, K. Synthesis of Solid-State NMR Study of MCM-41-type Mesoporous Silica
silica hollow nanoparticles templated by polymeric micelle with core− Nanoparticles. J. Am. Chem. Soc. 2005, 127, 3057−3068.
shell− corona structure. J. Am. Chem. Soc. 2007, 129, 1534−1535. (35) Apperley, D. C.; Harris, R. K.; Hodgkinson, P. Solid-state NMR:
(17) Sasidharan, M.; Zenibana, H.; Nandi, M.; Bhaumik, A.; Basic principles and practice; Momentum Press: 2012, DOI: 10.5643/
Nakashima, K. Synthesis of mesoporous hollow silica nanospheres 9781606503522.
using polymeric micelles as template and their application as a drug- (36) Martins, G. A. V.; Berlier, G.; Coluccia, S.; Pastore, H. O.;
delivery carrier. Dalton Trans. 2013, 42, 13381−13389. Superti, G. B.; Gatti, G.; Marchese, L. Revisiting the nature of the

14642 [Link]
Langmuir 2020, 36, 14633−14643
Langmuir [Link]/Langmuir Article

acidity in chabazite-related silicoaluminophosphates: combined FTIR

and 29Si MAS NMR study. J. Phys. Chem. C 2007, 111, 330−339.
(37) Poyraz, A. S.; Albayrak, C.; Dag, Ö . The effect of cationic
surfactant and some organic/inorganic additives on the morphology
of mesostructured silica templated by pluronics. Microporous
Mesoporous Mater. 2008, 115, 548−555.
(38) Kim, K.-M.; Kim, H. M.; Lee, W.-J.; Lee, C.-W.; Kim, T.-i.; Lee,
J.-K.; Jeong, J.; Paek, S.-M.; Oh, J.-M. Surface treatment of silica
nanoparticles for stable and charge-controlled colloidal silica. Int. J.
Nanomed. 2014, 9, 29.
(39) Fleming, I.; Williams, D., Infrared and Raman Spectra. In
Spectroscopic Methods in Organic Chemistry; Springer: 2019; pp. 85−
121.
(40) Cirtiu, C. M.; Raychoudhury, T.; Ghoshal, S.; Moores, A.
Systematic comparison of the size, surface characteristics and colloidal
stability of zero valent iron nanoparticles pre- and post-grafted with
common polymers. Colloids Surf., A 2011, 390, 95−104.
(41) Hammond-Pereira, E.; Bryant, K.; Graham, T. R.; Yang, C.;
Mergelsberg, S.; Wu, D.; Saunders, S. R. Mesoporous silica-
encapsulated gold core−shell nanoparticles for active solvent-free
benzyl alcohol oxidation. React. Chem. Eng. 2020, 5, 1939−1949.
(42) Reardon, E. J. Complexing of silica by iron (III) in natural
waters. Chem. Geol. 1979, 25, 339−345.
(43) Schindler, P. W.; Fürst, B.; Dick, R.; Wolf, P. U. Ligand
properties of surface silanol groups. I. Surface complex formation with
Fe3+, Cu2+, Cd2+, and Pb2+. J. Colloid Interface Sci. 1976, 55, 469−475.
(44) Pokrovski, G. S.; Schott, J.; Farges, F.; Hazemann, J.-L. Iron
(III)-silica interactions in aqueous solution: Insights from X-ray
absorption fine structure spectroscopy. Geochim. Cosmochim. Acta
2003, 67, 3559−3573.
(45) Su, Y.; Wang, Y.; Owoseni, O.; Zhang, Y.; Gamliel, D. P.; Valla,
J. A.; McPherson, G. L.; John, V. T. A Bottle-around-a-Ship Method
To Generate Hollow Thin-Shelled Particles Containing Encapsulated
Iron Species with Application to the Environmental Decontamination
of Chlorinated Compounds. ACS Appl. Mater. Interfaces 2018, 10,
13542−13551.

14643 [Link]
Langmuir 2020, 36, 14633−14643